SE524087C2 - A method for determining the friction between a surface and a tire for road vehicles driven with all wheels and a transmission clutch for distributing a torque between wheel axles comprising said method - Google Patents

Abstract

A system for determining friction between a surface and a tire of a driven wheel in a wheeled vehicle having a first and a second driven wheel, comprising the steps and functions of: - calculating slip values of a first and a second driven wheel dependent on the angular velocity of said first and second driven wheels; - calculating normalised traction forces of said first and second driven wheels dependent on the torque distribution between said first and second driven wheels; - calculating a friction estimate thetaf, thetar for each of said first and second driven wheels dependent on a predetermined relation between said calculated slip and said calculated normalised traction forces. The friction estimation system combined with a transmission coupling having a variable torque distribution.

Description

524 087 vehicles, there is no available wheel speed that is sufficiently reliable for the purpose of determining the available or actual friction between tire and road.

Object of the invention The general object of the present invention is to solve the problem of determining the available friction between a surface and a tire of a vehicle where there is no readily available vehicle speed.

One aspect of the problem is to determine such friction in an all-wheel drive road vehicle, especially when the vehicle speed is not directly divertable or derived from wheel speed signals and is thus unknown.

Summary of the Invention The above object is achieved and the problem is solved in the invention on a general level by combining information about the applied traction or torque and information about the resulting wheel sizing. More specifically, the solution is to combine slip information for at least two different drive wheels and information on traction derived by means of a pre-known distribution of torque between these wheels.

The slip information is preferably generated by wheel angular velocity information from wheel angular velocity sensors of each of said wheels. The torque distribution between the wheels is preferably known from a technical specification of the current all-wheel drive system of the vehicle. An estimate of the available friction between the surface and the tire is calculated in accordance with a predetermined model which is dependent on said wheel slip information and said torque distribution.

The invention is realized in various embodiments such as hardware, software or a combination thereof. The presently preferred embodiment is realized as a computer program product executed on a data processing system which takes input data from sensors of the vehicle.

Brief description of the drawings The invention will be explained in more detail with the aid of the accompanying drawings, in which: Figl shows slip curves for different road surfaces; .a. . va 1- n.: v:: z - H "II... u. .un ~.. _.. - eaa. - ø ~ o,.,,.... -.. - |... Fig. 2 - Fig. 5 shows plotted curves resulting from simulations of an embodiment of the invention, in which Fig. 2 shows torque and friction coils; Fig. 3 Fig. 4 shows estimated slip curve parameters for different constant torque distributions; Fig. 5 shows estimated slip curve parameters using a varying torque distribution; Fig. 6 shows a schematic block diagram / diagram of the functional invention and Figs. shows a schematic block diagram / fl fate diagram of road friction estimation combined with a transmission torque with variable torque distribution in a four-wheel drive (4WD) vehicle.

Detailed description of design features General concepts The design is based on the concept of wheel slip in relation to a road surface.

This relationship is usually presented as the longitudinal slip in percent as a function of the normalized traction applied to the wheels, resulting in the so-called slip curve. There are different slip curves for different surfaces, such as asphalt, gravel and ice for a road vehicle as shown in Fig. 1. More strictly speaking, the slip occurs between a tire of the wheel and a surface which is in contact with the tire. However, in this text, wheels and tires will be used as synonyms for the purposes of the invention in this regard.

The basic assumption is that the wheel slip s, defined as s = (a) -r - v) / v, and the normalized traction force, u are related to each other according to y = f (s). Here a) is the wheel angular velocity, r is the wheel radius and f (-) is a non-linear mapping. The interesting quantity to estimate for this purpose is the maximum value of f (-), i.e., p, = sup | f (s) I, which is the maximum available friction for the surface in question.

However, the lack of excitation, i.e. that only small to moderate slip values are available during normal running, in general that it is difficult to estimate the maximum value. The underlying approach used in the invention's road friction estimation is to estimate parameters that control the shape of the slip curve in an area around, u = 0 (eg slip factor and slip offset) using a parameterized model of the slip curve. The maximum value, thus corresponding to the available friction, is then calculated using the applied model and the resulting parameter estimates. 10 15 20 25 30 524 087 4 In the AWD case (all-wheel drive case), the problem aspect is to calculate a reliable estimate of the available friction without having reliable knowledge of <. ~. . | ~ - »now the vehicle speed. This is due to the fact that in the AWD case, all the wheels slip and therefore a calculation of the vehicle speed based on the angular speed of driving wheels gives an incorrect estimate of the true vehicle speed. See also the section on prior art above. The solution to this problem aspect and the idea behind this invention is to modify the underlying approach of manipulating all known relationships (ie equations) and variables to provide a reliable estimate of the available friction of an all-wheel drive vehicle.

In a general embodiment this is done by combining vehicle speed vi depending on slip equations y = f (s) for two different wheels, typically a front wheel and a rear wheel, and by replacing the vehicle speed vi one of the equations using an expression for v which has obtained by manipulating the other equations. This solves the problem of the unknown reference speed. However, a new problem aspect is introduced. More specifically, a new equation is obtained by substituting v in an equation (for example, that for the front wheel) with an expression for v obtained using the second equation (for the rear wheel). This new equation contains the non-analyzed traction forces for both wheels (or wheel axles) and not just the currently studied wheel (or wheel axle) for which the available friction is to be determined. However, in accordance with the invention, a model in the form of an equation system is specified depending on the torque distribution between the wheels (or wheel axles).

By utilizing a priori knowledge (prior knowledge) of this torque distribution, an estimate of the available friction is calculated from the system of equations. Prior art knowledge of torque distribution is generally known from technical specifications of modern commercially available AWD systems, such as Haldex LSC.

General Implementation Example In an embodiment of the invention applied to an AWD vehicle with a front wheel axle and a rear wheel axle, the general inventive concept or inventive concept summarized above is as follows.

Form the measured slimming via the expression (JJ-d) S____f r IN 60 V '10 15 20 524 087 5' where co, and w, are the angular velocities of the front f and rear r wheels respectively. Measured values for these angular velocities are preferably taken from existing wheel angle sensors in the vehicle. Let Y denote the total available torque after the gearbox. Then, assuming that the torque distribution between the front and rear axles is T f = yT T, = (1 - r) T where ye [OJ] is a known torque transmission parameter, the normalized tensile forces can be calculated as Tf HM = zffNf 'TI' " f ”= 241V, where r,, r ,, Ng, and N, are the wheel radii and the normal forces at the front axles and rear axles respectively.

By eliminating the unknown velocity at the slip equations for the front wheels and rear wheels, it can be shown that they are reduced to wr sf wff (s, + 1) -l. s, are related as sf "sf s + å III Ill s, + l where ö is the relative difference in radii between the front wheels and the rear wheels, ie å = V ', - I" f I "I' Furthermore, we also have u = f (S) QS = f "(u), so that the measured slimming and the normalized tensile forces are related to each other according to ff (u ,,,,.) +1 10 15 20 25 30 524 087 This can be reformulated as sm = g (uf, m 2 thin * ef 'Gr 7 1 where g (-) is a non-linear mapping and 9, and 9, are parameters that control the shapes of the front and back gluing curves near y = 0. This expression can be simplified further by the assumption that the tire road friction characteristics of the front and rear wheels are similar, this gives S .. = gwf ... - u ... ß), ie. it is only the difference in non-analyzed traction that matters. As before, it is assumed that 9 is dependent on the maximum friction, ie. ß = Hm.).

From these formulas and equations it appears that the quality of the rated 9 is critically dependent on the nature of the torque distribution y and is therefore used as an estimate of the maximum available friction. Note, however, that y = 1 corresponds to the front wheel drive drop (FWD = front wheel drive) while y = 0 corresponds to the rear wheel drive drop (RWD = rear wheel drive). If y = 0.5 constant then 9 cannot be estimated due to lack of identifi identbability, ie. that in this case there is no useful information in the signals that can be used to estimate the unknown parameter 9.

As y approaches the value 0.5, SNR tends to go towards zero, making it harder and harder to estimate the slip factor. One possible remedy in such a situation is to improve the level of excitation by requesting changes in the torque distribution. However, if y varies, as is the case in many modern AWD systems, 9 can be estimated, although the accuracy of the estimate will be best close to the front wheel drive (FWD) or rear wheel drive (RWD) situations, i.e. close to y = 1 and y = 0, respectively.

The friction parameter 9 can optionally be transferred to a suitable scale of values to adapt the friction estimate to presentation to a human user or for use in controlling other functional parts of the vehicle.

Detailed embodiment The parameterization g (-) used in the previous section is general and can be made arbitrarily complex. However, for practical use in an on-line implementation in a vehicle, this function should preferably have a simple form. A reasonable. . . . . »» - - n- 10 15 20 25 524 087 assumption is for the reason that for small slip values there is a linear relationship between slip s (t) and normalized traction force y (t), according to w) = fame, s, (t ) = p, (t) 6, where indices f and r indicate front axle and rear axle respectively. These terms can be used to form regression models for the unknown parameters 6 f and 9 ,. By further manipulating the shapes, it can be shown that the relationship between the true slimming and the measured slippage is S .. (f) ~ S, (f) -S, (f) +5 where 6 is the relative difference in radii between the front wheels and the rear wheels.

A straightforward parameterization that is suitable for estimation is therefore the model ef s., (F) = u) = (u, (f) - now) 1 ß, - ö Another perhaps more preferred approach is to assume that parameters 9 / and 6,, and thus the slip curves, are equal at both the front wheel and the rear wheel, i.e. k f = k ,. This gives the model mi) = Sf .. (f) = (HN) - u, (f) lie, which is general and also applies to front-wheel drive (FWD) and rear-wheel drive (RWD) vehicles.

The friction parameter 6 can be estimated both at the left and at the right side of the vehicle. This means a number of degrees of freedom in the formation of the final iction estimate. One approach is to consider the minimum value, ie. 6 = min (6,, 9,).

Another approach is to look at the mean, ie. 9 = (9, +6,) / 2.

Fig. 6 schematically shows in a combined block diagram and diagram the functions and steps of an embodiment of the invention. Thus, an angular velocity signal u) 312 including values for velocity signals for the angular velocity of drive wheels is input to the sampling stage 302. Optionally, the angular velocity signals are received in a digital form from an existing digital system. In step 304, slip values for drive wheels are calculated and output to step 306. In step 306, ... ... is calculated. . --_: _ - _ _ - _ _; _ ",. .-... _, - -. ß.. a n. - - zs 'u' ___ u... r.. - u .. fl - _ _ _.... a. v - - - _ 1 _ _ _ _ _......... - 10 15 20 25 30 524 087 analyzed traction forces of drive wheels in dependence on the calculated slip values in dependence on an input or pre-stored value 314 on In step 308a, a friction estimate is calculated depending on the calculated nonnalized traction forces and on a predetermined or pre-stored relationship between slimming and nonnalized traction forces.One or other friction estimation parameters from the side 308 are output, for direct presentation and on the other hand possibly to step 310 which occurs in embodiments of the invention In step 310 a setpoint value for torque distribution between drive shafts or wheels is generated by calculation or by selection from a pre-stored table, and a torque distribution signal is output for use in steering a v a transmission clutch with variable torque distribution between drive shafts. The setpoint is calculated or selected, for example, depending on the prevailing torque distribution in order to achieve a torque distribution which allows an accurate estimation of the available friction in the prevailing driving conditions. The setpoint for the torque distribution can also be calculated or selected depending on the friction estimate to achieve an improved driving situation with respect to sliding wheels.

An all-wheel drive (AWD) embodiment Fig. 7 shows an exemplary embodiment of an all-wheel drive vehicle (AWD vehicle) equipped with a transmission clutch with a variable torque distribution between drive shafts and drive wheels. Fig. 7 is a schematic sketch of a 4-wheel AWD vehicle 402 with drive wheels left front FL, right front FR, left rear RL and right rear RR. The vehicle 402 has a front transmission clutch 404 arranged to distribute torque between the drive wheels left front and right front, and a rear transmission clutch 406 arranged to distribute torque between the drive wheels left rear and right rear. The transmission clutch device is further provided with functionality to vary the torque distribution between the front and rear transmission clutches. In a variant, the front and rear transmission clutches are provided with functionality to vary the torque distribution between the left and right wheels.

An AWD transmission clutch suitable for the purposes of the invention with friction estimation combined with a transmission clutch that has variable torque distribution between driven wheel axles is designed, for example, in accordance with the Haldex LSC® transmission clutch. Such a construction comprises three functional main parts, more specifically a hydraulic pump driven in dependence on the slimming between the axles. .- .n n. n a .a s .a n ._ '. . n. o.. .n z: v z _ '_,., .. - s o.. _ _ _ 0 _ ".n. E. - a: Hv-I '_,. A.. -.... _ N 8....., ........ 10 15 20 25 30 524 087 »_ na / the wheels, a wet multi-plate clutch and a controllable throttle valve with an electronic control device. A piston actuator is connected to the other shaft.The two shafts are connected via the clutch package for the wet disc clutch, which is normally unloaded and thus does not transmit any torque between the shafts.When both shafts rotate at the same speed there is no pumping activity.Immediately when a speed difference occurs, the pumping is activated and starts to generate oil pressure.Because this is a piston pump, it is practically instantaneous reaction without any pumping loss due to low speed.The oil fl turns into a clutch piston which thus compresses the clutch package and brakes the speed difference m ellan axlarna. The oil returns to the reservoir via a controllable throttle valve, which controls the oil pressure and the force on the coupling package. When driving, ie. high slip conditions deliver a high pressure, while in tight curves, ie. when parking, or at high speeds, a much lower pressure is achieved. Again with reference to fi g. 7, wheel angular velocity sensors at each wheel supply wheel velocity signals co which are collected at an input stage 405 of a friction estimator 404. The friction estimator comprises in step 412 a road friction estimation functionality which receives as input angular velocity values and optionally controlled input data 408 technical parameters. The road friction estimation step 412 comprises friction estimation functionality in accordance with the embodiments described above and outputs a friction estimation in the form of a friction dependent parameter 6 410.

The friction parameter 9 410 is preferably available for presentation or use and is also input to step 414. In step 414 there is functionality for evaluating friction and torque distribution and for outputting a request signal 415 for a desired torque distribution depending on predetermined or dynamically input pairs.

The request signal 415 is input to a torque distribution control stage 416 which supplies a torque distribution control signal 418 to the transmission couplings 406, 404.

The transmission coupling device in turn varies the torque distribution in accordance with the control signal 418.

Simulation results To illustrate the performance and technical effect of the invention, this section describes the results of simulations of the estimation model applied below. n ..- »en u u ~ a n 40. _I <_ __. . . . . -. -: ß _ 1 _ _,,.

I 'fl' '"°' -. U» - n... - - -: - ~: °° '_ 2 _' '.. Q' 2 I '' °.... -... The use of a simple dynamic model together with the torque and friction problems shown in Fig. 2 has generated artificial data for the simulation. ~. .-.. 10 15 524 087 / O driving straight ahead with an AWD vehicle.

Start the simulation study by considering constant torque distributions, Figs. 3 shows the estimated 6 using the values y = 1 (corresponding to a front-wheel drive vehicle - FWD vehicle), y = 0.8 and y = 0.55. Plot F in g. 3 shows that the ability to estimate the 6 parameter decreases as the torque distribution y approaches the value 0.5. In the degenerate case where y E 0.5, the 6 parameter cannot be estimated at all due to lack of identifiability.

A perhaps more interesting simulation than the previous one is to assume that the torque distribution front / rear is a function of the actual slip. This is the principle used by most modern AWD systems. Fig. 4 shows the estimated 6 for an exemplary selection of such a function. The wheel gluing for the front wheels (upper curve) and rear wheels (lower curve) and the resulting torque distribution are shown in Fig. 5.

These simulations show that the friction parameter is accurately estimated by the invention as long as the torque distribution between front and rear wheel axles is not equal. However, as mentioned above, the identifiability can be improved by allowing the friction estimation algorithm to request changes in the y parameter.

. ~ -. -. - - I II 10 l5 20 25 524 087 Practical implementations A practically implemented embodiment of the invention preferably comprises computer program code parts comprising program code adapted to perform the mathematical operations described above. These operations are performed depending on input parameters from wheel angular velocity sensors and on a predetermined, possibly variable, torque distribution parameter which describes the torque distribution between a first and a second wheel or wheel axle.

The computer program code portions are preferably loaded into a digital data processing unit and adapted to control the data processing unit to perform the steps and functions of the invention. Values of input parameters are repeatedly sampled and processed to calculate and output an estimate of the prevailing available friction between the road and a specific wheel. This is done for all the wheels of the vehicle or for a selection of the wheels. The output estimate is used in various embodiments, for example for brake control, anti-slip control, alarms and the like.

It should be noted that the above embodiments are described as examples of the invention. For example, there are different combinations of first and second wheels and of parameters for first and second wheels in the calculations and equations. In the above examples, specific configurations of front and rear wheels are selected and presented, but it should thus be noted that other such configurations are conceivable. Other examples of such configurations are first wheel right front, second wheel left front; first wheel right rear, second wheel left rear; first rear wheels, second front wheels and so on.

Thus, the indices f and ri of the equations are generally interchangeable with indices first (1) and second (2).

Different applications The invention is advantageously applied in different types of road vehicles, but can also be applied in other wheeled vehicles such as robots or trucks. . .. Ü 'a w I I I II "" ° "" ° n: ao H -; ;; -, '. . . . . . . . . . .-

Claims (1)

1. 0 15 20 25 30 524 087 12 Patent claims. A method of determining friction between a surface and a tire of a drive wheel in a wheeled vehicle having a first and a second drive wheel, comprising the steps of: - calculating slip values of a first and a second drive wheel depending on the angular velocity of said first and second drive wheels; calculating normalized traction forces of said first and second drive wheels in dependence on the torque distribution between said first and second drive wheels; calculating a friction estimate 91, 62 for each of said first and second drive wheels depending on a predetermined ratio between said calculated slip and said calculated normalized traction forces. . The method of claim 1, wherein said slip values are calculated in a manner corresponding to the difference between the angular velocities of said first and second wheels, respectively, divided by the angular velocity of said second wheels. . The method according to any one of the preceding claims, wherein said slip values sm are calculated in a manner corresponding to the relationship _wf-w, m-90) where mf and co, are the angular velocities of the front and rear wheels, respectively. . The method according to any of the preceding claims, wherein said normalized tensile forces p p are calculated in a manner corresponding to the relationships T = f “f” zrNf _ T, “W zrN, where r, N f and N, are the wheel radius and the non-force at the front (f) respective rear wheel axle (s). . The method according to any one of the preceding claims, wherein said friction estimate is calculated by means of a relationship between the slimming and the normalized traction force beroende depending on the difference between the normalized traction forces of said first and second 10 15 20 25 10. 524 087. . ~. .- / 3 drivhjui. The method according to any one of the preceding claims, wherein said friction estimates are calculated by means of a linear relationship between sliming s (t) and non-analyzed traction force fi y (t) of said first and second drive wheels, respectively, corresponding to SfU) = Hf (l) 9f MI) = / 1, (f) 9, where indices f and r indicate front (f) and rear (r) wheel axles, respectively. The method according to the preceding claim, wherein the friction estimation pair frames 9f and 6 are determined by means of a regression model. The method according to any of the preceding claims, wherein the calculation of the slippage is further improved in a manner corresponding to the relationship between the true slimming sm and the measured slimming s fis, SMU) z SN) -S, (f) +5, where radii between the front wheel and the rear wheel, r, - rf ie. 6 =. The method according to any one of the preceding claims, wherein the estimate of the friction is calculated based on a model corresponding to ef sm (f) = s ”. (f) = (ti, (f) - m) 1 H,, 6 where ö is the relative difference in radii between the front wheels and the rear wheels, The method according to any of the preceding claims, it being assumed that the slip factor is the same in both said first and second drive wheels, and wherein the estimate of the friction is calculated based on a model corresponding to 10 15 20 25 30 ll. 12 13. 14. 15. 16. 17. 524 087 11+ SMU) = Sf ,,., (F) = (WU) _ 11,0) where ö is the relative difference in radii between the front wheels and the rear wheels, r, -rf ie.5 = I 'r The method according to any of the preceding claims, wherein the torque distribution changes when the torque distribution values are in the range of 0.5. . The method according to any of the preceding claims, wherein a value for a desired torque distribution is generated in dependence on the prevailing torque distribution between said first and second drive wheels. The method according to any one of the preceding claims, wherein a value for a desired torque distribution is generated in dependence on the prevailing available friction for said first and second drive wheels. The method according to any of the preceding claims, wherein a value for a desired torque distribution is generated and input to a mechanism for varying torque distribution in a transmission clutch with a variable torque distribution. The method according to any of the preceding claims, wherein said first and second drive wheels are front wheels and rear wheels, respectively, in a four-wheel drive vehicle. An apparatus for determining friction between a surface and a tire of a drive wheel in a wheeled vehicle having a first and a second drive wheel, comprising means for performing the steps and functions of any of the preceding claims. A computer program product for determining friction between a surface and a tire of a drive wheel in a wheeled vehicle having a first and a second drive wheel, comprising computer program code portions adapted to control a data processing system to perform the steps and functions of any of the preceding claims. 52 I Û u .... a. .n ø _ '° _ _ ",. n .. u a. en v ** _ _. .. .. .. ...: _ _ _ .....- -3; -; ~ ~, _, Is '' 1 ". . - - ~ - ~ - ~ ~ - “18. A transmission clutch for distributing a torque between the first and second 10 15 20 25 30 19. 20. 21. wheel axles of a wheeled vehicle having a first and a second drive wheel, the transmission clutch having a mechanism for varying the distribution of torque between said first and second wheel axles and a steering mechanism for controlling said variation of said torque distribution, said steering mechanism comprising: - means for determining friction between a surface and a tire of a drive wheel in said wheeled vehicle by means for calculating the slip values of a first and a second drive wheel in dependence on the angular velocity of said first and second drive wheels; means for calculating normalized traction forces of said first and second drive wheels in dependence on the torque distribution between said first and second drive wheels; means for calculating a friction estimate 9,, 62 for each of said first and second drive wheels in dependence on a predetermined relationship between said calculated sliming and said calculated non-analyzed traction forces; and - means for evaluating said friction and a prevailing torque distribution between said first and second wheel axles. The transmission clutch according to the preceding claim, further comprising means for generating a control signal for torque distribution in dependence on said friction and said prevailing torque distribution. The transmission clutch according to the preceding claim, wherein the control signal for torque distribution is input to the mechanism for varying torque distribution. The transmission coupling according to any one of the preceding claims 18 to 20, further comprising means for performing the steps and functions according to any one of claims 1 to 17.